Statistical models fed by massive data that no single human can fully understand are still “science”, but it’s definitely a very different paradigm for science

Statistical models fed by massive data that no single human can fully understand are still “science”, but it’s definitely a very different paradigm for science than what we’ve had for the past few centuries, and it seems to be what we’re heading towards. Eg a few decades from now, 24/7 biometrics monitoring at massive scale + machine learning might let us predict complex conditions like Alzheimer’s, Parkinson’s, various types of cancer, etc very accurately and very early, yet we might not be any closer to understanding what causes them at a fundamental level in the first place.